Side Lesson 29: Currency Hedging — When to Hedge, How to Hedge, and the Cost of Insurance

---

1. Why This Is Important

Side lesson 18 told you the punchline first: this course's recommended portfolio is U.S.-listed equities only. So why is there a whole side lesson on currency hedging? Because **the U.S.-only rule is about equities, not bonds, and it has exceptions.** The moment you reach outside the United States for anything — an EAFE sleeve, an international bond fund, a foreign-currency cash deposit, a Swiss holding company that owns a U.S. ADR — currency arithmetic is back in your portfolio whether you wanted it or not.

There are four reasons this is worth a full side lesson rather than a footnote:

**The "leave international unhedged" reflex is the wrong default

for international bonds.** For stocks the conventional wisdom is roughly defensible (currency vol ~7-9%/yr is small relative to equity vol ~16%/yr, and over 10+ years the FX leg has roughly zero expected return). For investment-grade ex-U.S. bonds the currency vol exceeds the asset vol — a 4% yielding German bund with 8% currency volatility is a worse Sharpe ratio than a U.S. T-note. If you own non-U.S. bonds and you don't hedge, you have accidentally turned a fixed-income sleeve into an FX trade.

Hedging cost is not trivia, it is a measurable price.

Covered-interest-parity makes the cost of a one-year currency hedge approximately equal to the interest rate differential. When U.S. T-bills yield 4.3% and Japanese T-bills yield 0.3%, the one-year USD/JPY hedge earns you about 4 percentage points; when euro rates briefly exceeded U.S. rates in 2008-2009, the hedge cost you ~50bp. Knowing this number turns "should I hedge" into a calculation, not a vibe.

Hedged ETFs exist and are cheap. HEFA (iShares EAFE hedged)

charges 35bp; HEDJ (Europe hedged) and HEWJ (Japan hedged) are in the same range. The hedge mechanic — rolling one-month forwards — is not exotic, it is a baked-in feature you can buy or skip with one ticker change. The retail toolkit is fully built out.

The dollar regime moves in long swings, not noise. The DXY

spent 2002-2008 falling, 2008-2014 ranging, 2014-2022 rising sharply, and 2024-2026 retracing. These are six- to ten-year regimes — long enough that being wrong-footed once costs you a chunk of an investing lifetime. You do not have to predict the next regime to size around the last three.

This lesson is not telling you to hedge. It is telling you what the trade-off looks like, where the corner cases are, and which ETFs to use when you decide it is worth doing.

---

2. What You Need to Know

2.1 The Two-Bet Decomposition

Every dollar you put into a foreign asset is two bets:

The local-currency return on the asset — the stock or bond

priced in its home currency.

The currency return — the change in exchange rate between

that home currency and the U.S. dollar over the holding period.

The exact dollar-denominated return is the product:

$$ 1 + R_{\text{USD}} \;=\; (1 + R_{\text{local}}) \times (1 + R_{\text{FX}}) $$

For small returns the cross-product is negligible and the decomposition is approximately additive: $R_{\text{USD}} \approx R_{\text{local}} + R_{\text{FX}}$. A worked example: in 2024 the Nikkei 225 returned about +19.2% in yen terms; the yen depreciated about 11% against the dollar; the unhedged USD return was therefore $(1.192)(0.89) - 1 \approx +6.1\%$. A Japanese investor saw +19%, an American who held the same stock unhedged saw +6%, and an American who held it hedged saw approximately +19% minus the hedge cost.

The volatility decomposition under the assumption of independent FX and local returns is:

$$ \sigma^2_{\text{USD}} \;=\; \sigma^2_{\text{local}} + \sigma^2_{\text{FX}} $$

For an unhedged developed-market equity, this is roughly $16^2 + 8^2 = 17.9\%$ — about 12% higher portfolio vol than the local. For an unhedged developed-market 7-year bond it is $5^2 + 8^2 = 9.4\%$ — vol is double the underlying. The FX leg matters categorically more for bonds than for stocks.

2.2 What a Hedged ETF Actually Does

A "currency-hedged" ETF holds the foreign basket and sells a series of one-month foreign-currency forwards equal in notional to the basket. At each month-end the forward expires, the realised FX P&L is settled in dollars, and a new forward is opened. The mechanical effect: for the one month between rolls, the dollar return tracks the local-currency return almost exactly, less a small drag from forward pricing.

The drag is the interest-rate differential, which by covered-interest-parity equals the percentage difference between the spot rate and the forward rate:

$$ \text{Forward}/\text{Spot} \;\approx\; \frac{1 + r_{\text{foreign}}} {1 + r_{\text{US}}} $$

Re-arranged for an annualised hedge cost:

$$ \text{Hedge cost (per year)} \;\approx\; r_{\text{US}} - r_{\text{foreign}} $$

This is a sign-honest number. If U.S. short-term rates exceed foreign rates (the situation in April 2026: U.S. 4.3%, EUR 2.0%, JPY 0.5%, GBP 4.0%), the hedge earns positive carry of ~2.3% against the euro and ~3.8% against the yen before the ETF's expense ratio. If the spread inverts (the U.S. ZIRP era 2009-2015 against most of the world), the hedge bleeds carry — but this is rare and shallow, typically 0-50bp against major currencies.

Three vehicles to know:

• HEFA — iShares Currency Hedged MSCI EAFE. 35bp ER. AUM ~$8B.

The hedged twin of EFA. Pairs cleanly with the unhedged sibling for comparison.

• HEDJ — WisdomTree Europe Hedged Equity. 58bp ER. AUM ~$2B.

Tilts toward export-heavy Europe (Siemens, ASML, LVMH, Sanofi).

• HEWJ — iShares Currency Hedged MSCI Japan. 35bp ER. AUM ~$0.4B.

The hedged twin of EWJ.

The 35bp ER is on top of the underlying basket's natural cost. The hedge mechanic itself adds approximately 5-10bp of frictional cost (monthly roll, bid-ask on forwards) on top of any carry signal.

2.3 The Long-Run Equivalence Theorem

A robust empirical regularity: over rolling 10- to 20-year windows, **hedged and unhedged international equity returns are within roughly 10-30bp/yr of each other.** The mechanism is straightforward — over long enough horizons real exchange rates revert to purchasing power parity, so the cumulative FX contribution to the unhedged return averages roughly zero, leaving you with just the asset return either way.

What does not equalise: volatility and drawdowns. The hedged sleeve has consistently lower realised vol (about 16% vs 18% on EAFE since 2002) and shallower max drawdowns. So the long-run trade-off collapses to: same expected return, lower vol — i.e., a higher Sharpe ratio for the hedged sleeve. This is the rare case where there is a free lunch in the data, and the cost of getting it is just the 35bp ER.

The reason most retail advisors still recommend "leave international unhedged" is behavioural rather than mathematical. Holding HEFA against a falling dollar means underperforming the headline EFA in a dollar-bear regime, and clients fire advisors who underperform the benchmark even when total-return Sharpe is better. Institutional money that does not have that constraint hedges roughly 50-75% of its ex-U.S. equity exposure.

2.4 Short-Run: Five Regimes That Mattered

Sub-decade windows are where the hedge/unhedge decision actually shows up in your statement. The post-1990 record:

• 1995-2002 — strong dollar. DXY ran from 80 to 120. Unhedged

ex-U.S. bled relative to hedged. Net cost of being unhedged: ~3-4% per year for seven years.

• 2002-2008 — weak dollar. DXY dropped from 120 to 71. Unhedged

won by ~3-4% per year. This is the regime that locked in the textbook recommendation to leave international unhedged.

• 2014-2016 — strong dollar 2.0. DXY surged from 80 to 100 over

18 months as the Fed taper diverged from ECB QE. HEFA outperformed EFA by about 9 percentage points cumulative.

• 2017 — weak dollar. DXY fell 10% as Trump-tax-cut optimism

faded. Unhedged EFA returned 25%; HEFA returned 17%. The unhedged sleeve won by 8 percentage points in a single year.

• 2022 — strong dollar 3.0. DXY hit 114, the highest since 2002,

as the Fed hiked 525bp into a global recession. HEFA -7.8%; EFA -16.8%. The hedge saved 9 points of drawdown.

• 2024-2026 — dollar retracement. DXY drifted from 107 to 99.

Unhedged EFA modestly outperformed. The current April-2026 setup is the kind of regime where neither side has a clear edge.

The pattern: regimes are long (5-10 years), the magnitudes within them are large (3-4%/yr cumulative), and they do not telegraph in advance. That is what makes hedging a real allocation question rather than a pure "always" or "never."

2.5 The U.S.-Only Carve-Out: International Bonds

The U.S.-only equity rule from side lesson 18 has one quietly important exception: **if you hold non-U.S. fixed income at all, you should hedge it.**

The math from §2.1: a 4-5% yielding ex-U.S. investment-grade bond sleeve has roughly 5% local-price volatility and 8% currency volatility. Combined, the unhedged sleeve has more volatility than the bond's running yield — meaning the FX leg dominates. You are holding a vehicle that the marketing material calls "fixed income" but that behaves like an FX trade with a coupon strapped to it.

Hedged international bond ETFs:

• BNDX — Vanguard Total International Bond. 7bp ER. ~$50B AUM.

Hedged USD. The default if you want any non-U.S. fixed income at all.

• IAGG — iShares Core International Aggregate Bond. 7bp ER.

Hedged USD. BNDX's main competitor.

Note both are hedged by default at the product level — there are no significant unhedged international IG bond ETFs trading at meaningful AUM, because the buyers of these products are the institutions that already did the math in this section.

The course's actual position: own VGIT / IEF / TLT for duration, and own BNDX only if you want explicit non-U.S. credit/duration diversification. For most retail portfolios, the answer is "skip BNDX entirely; the diversification benefit is small once both vehicles are USD-hedged duration." But if you do reach for it, do not hold the unhedged version. There is no scenario in which an unhedged ex-U.S. IG bond sleeve is the optimal choice for a USD-spending investor.

2.6 Optimal Hedge Ratio: The Decision Tree

You do not have to choose 0% hedged or 100% hedged. The hedge ratio $h$ that minimises portfolio volatility is given by the regression coefficient of unhedged return on FX return — which has a closed form:

$$ h^* \;=\; 1 + \rho \cdot \frac{\sigma_{\text{local}}}{\sigma_{\text{FX}}} $$

where $\rho$ is the correlation between the local-currency asset return and the FX return. For developed-market equities $\rho$ is typically slightly positive (about +0.1 to +0.2), so $h^*$ comes out a little above 1.0 — fully hedge. For commodities and EM equities $\rho$ is typically negative (the local market sells off when its currency weakens), so $h^*$ is well below 1.0 — partial hedge or skip.

The pragmatic decision rule the course uses:

Sleeve	Recommended hedge ratio	Vehicle
U.S.-listed equity	n/a	VTI/SPY
U.S.-listed ADRs	0%	TSM, ASML, etc.
Ex-U.S. IG bonds	100%	BNDX, IAGG
Ex-U.S. developed-market equity	50-100%	HEFA + EFA mix
Emerging-market equity	0-50%	EEM unhedged or HDEM partial
Commodities (priced in USD)	0%	DBC, PDBC
Gold (USD price)	0%	GLD, IAU

The 50-100% ex-U.S. equity range is wide on purpose. It reflects that inside this course's framework you should not have a large ex-U.S. sleeve at all — and if you do, the choice between 50/50 and 100/0 is mostly about behavioural tracking-error tolerance, not about expected returns.

---

3. Common Misconceptions

"Hedging is expensive." When U.S. rates exceed foreign rates

(the modal regime since 2008), hedging earns positive carry. The "expense" is just the ETF's own ~35bp expense ratio.

"Hedging is gambling on the dollar." It is the opposite.

Not hedging is taking a position on the dollar. Hedging removes the FX bet so you only own the underlying asset.

"Forward rates predict future spot rates." They do not. The

forward rate equals the interest-rate differential by covered-interest-parity. Empirically, the spot rate at expiration has zero correlation to the forward — the "forward premium puzzle."

"Currency volatility averages out over 10 years." The

cumulative return contribution averages to roughly zero over 10+ years. The volatility contribution does not — it adds to your monthly drawdowns the entire time.

"International bonds give you currency diversification." They

give you currency risk. If you wanted currency diversification you would buy an FX product directly, not bond-FX combo.

"Hedged ETFs are leveraged." They are not. The forward leg is

fully collateralised by the underlying basket. Counterparty risk exists but is a fraction of a percent of NAV.

"DXY measures the dollar against everything." It is a

trade-weighted index of six currencies (EUR 57.6%, JPY 13.6%, GBP 11.9%, CAD 9.1%, SEK 4.2%, CHF 3.6%). The Fed's broader DTWEXBGS includes 26 trading partners and is what hedging cost actually tracks.

"If I'm a long-term investor I don't need to hedge." Only true

for ex-U.S. equities, and only because the asset vol dominates the FX vol on long horizons. False for ex-U.S. bonds, where FX vol is the larger component at every horizon.

"Hedging removes the interest-rate-differential pickup." It

does the opposite. The differential becomes the hedge carry. You cannot collect a higher foreign yield without taking the FX exposure that on average kills it.

"Just buy a few foreign stocks directly to keep it simple."

You will under-diversify, pay foreign withholding tax, and have no easy way to layer a hedge. Use a U.S.-listed ETF (hedged or unhedged) for any non-U.S. exposure.

---

4. Q&A Section

**Q1: I own VXUS in my 401(k). Should I switch to a hedged alternative?** For an equity sleeve held 10+ years the long-run expected returns of hedged vs unhedged are nearly identical. If your 401(k) offers a hedged developed-market option at <50bp ER, the Sharpe ratio is modestly better hedged. If it does not, leaving VXUS alone is fine — the inferior option is "sell to chase the recent winner." The course position remains: minimise the ex-U.S. equity sleeve in the first place.

**Q2: U.S. rates are 4.3% and Japanese rates are 0.5%. Is now a great time to hedge yen exposure?** Yes, in a narrow sense — the carry is roughly +3.8%/yr in your favour. But the carry is already priced into the forward rate, which is what the hedged ETF buys. The carry shows up as the hedge ratio mechanically working in your favour each month, and is the reason HEWJ has tracked EWJ-in-yen so closely since the BOJ kept rates near zero.

Q3: What is the difference between DTWEXBGS and DXY? DXY is the ICE Dollar Index, six currencies, ~58% euro by weight, launched 1973. DTWEXBGS is the Fed's Trade-Weighted Broad Goods + Services index, 26 currencies including CNY and MXN, recalibrated annually. DTWEXBGS is the better measure of the dollar's economic strength; DXY is the better measure of what financial markets watch (and what hedged-EAFE ETF managers actually trade).

Q4: Can I hedge currency exposure myself with futures? You can, but it is not worth it. /6E (euro futures) is $125k notional per contract; rolling quarterly produces tracking error and has tax treatment under §1256 (60/40 LTCG/STCG, see week 39). For sleeves under $5-10M, the 35bp HEFA expense ratio is cheaper than the operational drag of doing it yourself.

**Q5: Why do hedged ETFs sometimes underperform their unhedged twin even when the dollar is rising?** Three reasons: (1) the hedge resets monthly, so intra-month FX moves inside the dollar's overall trend can show up wrong-footed; (2) the ETF's expense ratio is 30-40bp higher than the unhedged version; (3) the forward roll incurs small bid-ask costs.

Q6: The course says U.S.-only. Why are you teaching me how to hedge? The U.S.-only rule is about equity recommendations. The carve-outs are (a) U.S.-listed ADRs (already in U.S. dollars, no FX exposure to hedge); (b) international IG bonds if you choose to own them, where you should always hedge; (c) the rare investor who insists on a 10-20% ex-U.S. equity sleeve, where the hedge decision matters. The lesson exists for completeness, not as a recommendation to add ex-U.S. exposure.

Q7: What about emerging-market currencies — INR, BRL, ZAR? Most major EM currencies do not have liquid forward markets at retail sizes. EM-equity ETFs like EEM are unhedged by structural necessity. The few hedged EM products (HEEM was de-listed in 2018) failed because EM currency carry against USD is positive — i.e., hedging costs you 3-5%/yr of yield, and the EM equity correlation with EM FX is strongly negative, meaning the unhedged position partly diversifies itself.

Q8: Does Berkshire Hathaway hedge its foreign-currency exposure? Famously almost never. Buffett's view is that PPP (purchasing power parity) reverts on long horizons and the hedge cost over BRK's multi-decade holding period is therefore a deadweight loss. This is defensible at BRK's scale and time horizon. It is not defensible for a retail investor with a 10-year horizon and a finite tolerance for intra-decade drawdowns.

Q9: What happens to my hedge when there is a currency crisis? The forward leg pays out in dollars at maturity even if the foreign currency has depreciated 50%. Counterparty risk is small (forwards are collateralised at major dealers). The bigger risk is that during the crisis the underlying foreign asset also crashes — the 1998 Asia crisis hit EM bonds (-30%) regardless of hedge status, because the local-currency leg was the disaster.

Q10: Why doesn't covered-interest-parity break in extreme stress? It does. In late 2008 USD funding stress widened the deviation from CIP to several hundred basis points (the "USD basis"). It does not matter for retail because the hedged ETF's market price reflects the strain in real-time and you can sell out at NAV without rolling forwards yourself. It matters for prime brokers running cross-currency repo books, which is not your problem.

Q11: How does §1256 tax treatment apply to currency hedges? For the futures used by some hedged ETFs internally, yes — but the ETF wrapper handles all of it. Your 1099 from a hedged ETF reports ordinary distributions and capital gains the same as any other ETF. You do not see the §1256 mechanics directly. (See week 39 for the direct-futures version of the same trade.)

**Q12: Is there a "hedged S&P 500" for non-USD investors that I should know about?** Yes — IWDA-hedged variants exist on European exchanges, and similar USD-hedged S&P products exist for Asian investors who want U.S. equities translated back to home currency. Mirror image of HEFA. Not relevant if you are a U.S.-domiciled investor (your spending is in USD), but useful to know if you are advising a non-U.S. friend who holds U.S. stocks.